[论文解读] Aortic Valve Disease Detection from PPG via Physiology-Informed Self-Supervised Learning

Jiaze Wang, Qinghao Zhao+5 more|arXiv (Cornell University)|Feb 4, 2026

Cardiac Valve Diseases and Treatments被引用 0

一句话总结

本文提出 Physiology-Guided Self-Supervised Learning (PG-SSL)，利用无标签 PPG 数据筛查 AS/AR，在 AUC 上优于监督基线，具备强的预后价值。

ABSTRACT

Traditional diagnosis of aortic valve disease relies on echocardiography, but its cost and required expertise limit its use in large-scale early screening. Photoplethysmography (PPG) has emerged as a promising screening modality due to its widespread availability in wearable devices and its ability to reflect underlying hemodynamic dynamics. However, the extreme scarcity of gold-standard labeled PPG data severely constrains the effectiveness of data-driven approaches. To address this challenge, we propose and validate a new paradigm, Physiology-Guided Self-Supervised Learning (PG-SSL), aimed at unlocking the value of large-scale unlabeled PPG data for efficient screening of Aortic Stenosis (AS) and Aortic Regurgitation (AR). Using over 170,000 unlabeled PPG samples from the UK Biobank, we formalize clinical knowledge into a set of PPG morphological phenotypes and construct a pulse pattern recognition proxy task for self-supervised pre-training. A dual-branch, gated-fusion architecture is then employed for efficient fine-tuning on a small labeled subset. The proposed PG-SSL framework achieves AUCs of 0.765 and 0.776 for AS and AR screening, respectively, significantly outperforming supervised baselines trained on limited labeled data. Multivariable analysis further validates the model output as an independent digital biomarker with sustained prognostic value after adjustment for standard clinical risk factors. This study demonstrates that PG-SSL provides an effective, domain knowledge-driven solution to label scarcity in medical artificial intelligence and shows strong potential for enabling low-cost, large-scale early screening of aortic valve disease.

研究动机与目标

• 推动在超声心动图限制之外实现可扩展、低成本的 AVD 筛查。
• 利用大规模无标签 PPG 数据学习与疾病相关的形态学特征，而无需直接标签。
• 将 AS/AR 的生理知识转化为代理自监督任务用于预训练。
• 在少量带标注的 AS/AR 子集上以高效的双分支架构进行微调。
• 证明其预后价值并与标准临床风险因素独立。

提出的方法

• 将临床 AVD 生理学形式化为可计算的 PPG 形态学表型，以创建代理前任务。
• 在 UK Biobank 的 170,702 个无标签 PPG 样本上进行预训练，以学习领域特定特征。
• 使用双分支门控融合架构，在少量带标注的 AS/AR 子集上进行微调。
• 将 PG-SSL 与监督基线和通用 SSL 方法（SimCLR、重建、K-means）进行比较。
• 利用 Grad-CAM 可视化模型关注区域，以验证特征的生理相关性。
• 通过多变量 Cox 回归和倾向性分数匹配评估其与临床风险因素的独立预后性。

Figure 1 : Performance evaluation of the model on the independent test set. (A, B) ROC curves for AS and AR detection. The model achieves an AUC of 0.765 for AS and 0.776 for AR. (C) Calibration curves. The axes focus on the [0, 0.45] interval, showing high consistency between predicted probabilitie

实验结果

研究问题

• RQ1 无标签 PPG 上的生理信息自监督预训练能否在标注数据有限的情况下提升 AS/AR 筛查效果？
• RQ2 学到的特征是否提供超越标准临床风险因素的独立预后价值？
• RQ3 与通用 SSL 方法及传统监督模型相比，PG-SSL 在 AS/AR 任务中的表现如何？
• RQ4 模型注意力是否与已知的 AS 和 AR 的血流动力学机制对齐？
• RQ5 是否可以利用早期诊断时间窗来提高筛查性能？

主要发现

• PG-SSL 在独立测试集上对 AS 的 AUROC 为 0.765，对 AR 的 AUROC 为 0.776。
• 在 60% 特异性下，AS 的 S@60Sp 为 77.6%，AR 的为 78.6%。
• PG-SSL 优于监督基线和若干 SSL 基线（如 SimCLR、重建）。
• 在 Cox 模型中，模型预测在调整临床风险因素后仍具备预后意义。
• Grad-CAM 可视化显示的关注模式与 AS 与 AR 的生理特征相一致，具有生理学上的合理性。
• 手工特征与深度特征融合后性能下降，表明 PiLA 特征具高度完备性。

Figure 2 : Grad-CAM visualization of model attention across different groups. (Left) Healthy controls show diffuse attention, verifying holistic waveform integrity. (Center) AS patients show focused attention on the delayed systolic upstroke and dicrotic notch. (Right) AR patients exhibit broad cove